[R] using sampling weights in glm

[Thomas Lumley]

On Wed, 15 Aug 2007, Werner Wernersen wrote:

> Hi,
>
> I have a supposedly representative sample from a
> stratified survey. Hence, all observations have an
> associated sample weight each which inflate the sample
> to population size.
>
> I want to run a glm probit regression on the data but
> I am not clear about if or better how I can use the
> weights in it. From the description of the weights
> argument of glm it seems to me that I cannot plug
> these weights in there.
>

You want svyglm() in the survey package.

      -thomas


Thomas Lumley			Assoc. Professor, Biostatistics
tlumley at u.washington.edu	University of Washington, Seattle